×

You are using an outdated browser Internet Explorer. It does not support some functions of the site.

Recommend that you install one of the following browsers: Firefox, Opera or Chrome.

Contacts:

+7 961 270-60-01
ivdon3@bk.ru

  • Methods for determining the target sentence for automated test questions generation

    Automation of test question generation consists of four main stages. The first step is to determine the target sentence from which it will be possible to generate a test question. This article examines the existing methods for determining the target sentence Both foreign and Russian language sources are considered. The paper considers such methods as automatic text abstracting, similarity of syntactic analysis, a rule-based method, a method based on ranking by keywords, an algorithm for summarizing documents based on the frequency of words.

    Keywords: automatic test questions generation, automatic text processing, natural language processing, automatic abstracting, annotation

  • Methods for recognizing named entities in Russian

    This article briefly discusses recurrent neural networks and bidirectional recurrent neural networks. A conditional random field is also described. In particular, we consider the vector representation in the form of the ELMo model, the bidirectional language model and its fine-tuning. A diagram of the architecture of a bidirectional recurrent neural network with CRF is also given.

    Keywords: named entity recognition, neural network, word vector representation, language model, conditional random field, neural network architecture, recurrent neural network, natural language processing, bidirectional recurrent neural network

  • Algorithms for the chatbot to search for products

    This article discusses the algorithms used for the operation of the chatbot. Within the framework of these algorithms, a general approach to the operation of dialog systems is considered. And the algorithm of the dialog system is proposed, based on a general approach, which includes a natural language processing module. The algorithms that will be applied to the chatbot to search for products of interest to the user are also presented. Among the considered algorithms are such as the algorithm of interaction between the user and the chatbot, the processing of user requests and the algorithm of natural language processing. A modified algorithm for natural language processing is also presented, taking into account the necessary requirements.

    Keywords: chatbot, natural language processing, user request processing, named entity recognition, dialog manager, dialog system